TK.

Ok, I found the error: it doesn't work with two 32k BankSticks since there is a test which ensures that a 64k BankStick is connected before formating the song partition. This function should not be called on two different 32k BankSticks of course. I tested it with two 24LC512 - and this test passed, therefore I didn't know about this issue. However, you will find a patch under http://www.ucapps.de/tmp/mbseq_v2_0_bs_fix.zip - it will be part of the next release. Best Regards, Thorsten.

Hi, the result of #if LCD_USE_SECOND and #if LCD_USE_SECOND == 1 is the same, so this cannot be the reason. The only changes of the modified lc_mproc.inc and the original lc_mproc.inc of v1.4 are: [avatar@britannia mblc_sonar_mod]$ diff -c /mnt/data/mios_apps/release/midibox_lc_v1_4/lc_mproc.inc lc_mproc.inc *** /mnt/data/mios_apps/release/midibox_lc_v1_4/lc_mproc.inc 2003-09-29 14:22:28.000000000 +0200 --- lc_mproc.inc 2004-03-29 01:43:07.000000000 +0200 *************** *** 360,365 **** --- 360,369 ---- bz LC_MPROC_Action_Query movf LC_MPROC_ACTION, W + xorlw 0x1a + bz LC_MPROC_Action_MCQuery + + movf LC_MPROC_ACTION, W xorlw 0x02 bz LC_MPROC_Action_HostReply *************** *** 373,379 **** movf LC_MPROC_ACTION, W xorlw 0x12 ! bz LC_MPROC_Action_WriteLCD movf LC_MPROC_ACTION, W xorlw 0x13 --- 377,384 ---- movf LC_MPROC_ACTION, W xorlw 0x12 ! skpnz ! rgoto LC_MPROC_Action_WriteLCD movf LC_MPROC_ACTION, W xorlw 0x13 *************** *** 427,432 **** --- 432,463 ---- rgoto LC_MPROC_ActionFinished ;; -------------------------------------------------------------------------- + ;; MIDI Action: MC Query (undocumented) + ;; -------------------------------------------------------------------------- + LC_MPROC_Action_MCQuery + ;; wait until sysex footer (F7) has been received + movlw 0xf7 + IFNEQ LC_MPROC_IN, ACCESS, rgoto LC_MPROC_SysExCheck_End + + ;; send SysEx header + call LC_MPROC_Send_SysExHeader + + ;; send reply + movlw 0x1b + call MIOS_MIDI_TxBufferPut + + ;; send dummy byte (seems to work) + movlw 0x42 + call MIOS_MIDI_TxBufferPut + + ;; send of SysEx footer + movlw 0x01 ; (independend from merger state) + call LC_MPROC_Send_SysExFooter + + ;; finish Action + rgoto LC_MPROC_ActionFinished + + ;; -------------------------------------------------------------------------- ;; MIDI Action: LC Host Reply ;; -------------------------------------------------------------------------- LC_MPROC_Action_HostReply so: the new LC_MPROC_Action_MCQuery and a change for the LC_MPROC_Action_WriteLCD branch (distance between the "bz" and the function was > 128 bytes, therefore a rgoto has to be used). It should work if you just overwrite the old lc_mproc.inc with the modified one. Maybe you've modified the source code in a way which prevents the LC_MPROC_Action_WriteLCD function from being called... Best Regards, Thorsten.

I guess that a microcontroller based vu meter would require a seperate core module + a lot of external electronic to prepare the audio signal and especially to convert the db scaling + the appr. software. So, it's simpler to use a dedicated IC for this job like the LM3916. See this project http://www.all-electric.com/schematic/quad-met.htm Best Regards, Thorsten.

Hallo, danke fuer das Lob - es motiviert zu weiteren Taten :) Das Poti wird zwischen 0V und 5V geschaltet, die Spannung am Schieber liegt also immer zwischen diesen beiden Werten, unabhaengig vom Widerstandswert. Jedoch: je kleiner der Wert, desto hoeher der Stromverbrauch, je groesser der Widerstand, desto hoeher der Rauschanteil. Deshalb sollte der Poti-Wert zwischen 1k und 10k liegen (Erfahrungswerte) - 100k sind definitiv zu viel. Und der Spannungsverlauf sollte linear sein. Beim Taster spielen die Spannungs/Stromwerte keine Rolle, der Stromverbrauch ist minimal. Gruss, Thorsten.

Ja, wenn MIOS bereits drauf ist, sollte das LCD etwas anzeigen. Gruss, Thorsten.

Theoretisch koennte man Gruppen ueber die MB64E_ENC_UpdateMap Funktion in mb64e_enc.inc realisieren. Hier werden die Encoder momentan je nach selektierter globaler Gruppe auf die entsprechenden Encoder Eintraege gemapped. Dich man koennte natuerlich auch weitere Gruppen definieren... Dafuer stehen allerdings "nur" 128 Encoder Eintraege zur Verfuegung (oder wie ein Amerikaner schreiben wuerde: es steht die absolut gigantische Menge von bis zu 128 Encoder Eintraegen zur Verfuegung, die bei eingehenden MIDI Events sogar gleichzeitig verarbeitet werden koennen - das kann bisher kein anderer MIDI Controller auf dem Markt ;-)). Mehr geht nicht, dafuer ist das MB64E Konzept nicht ausgelegt. Braucht aber auch keiner - denn das Mapping kann man genausogut ueber die Host Software realisieren. Nuendo sollte eigentlich genauso wie Cubase SX in der Lage sein, mit selbstdefinierten Controller Gruppen umzugehen, die sich dann via MIDI umschalten lassen. Auf diese Weise koenntest Du beliebig viele Gruppen definieren, die dann von der Host Software verwaltet werden - oder nicht? Jedenfalls waere das die geeignetere vorgehensweise Gruss, Thorsten.

How does MBSEQ behave with: #define DEFAULT_BANKSTICK_PATTERN_CS 1 #define DEFAULT_BANKSTICK_SONG_CS 0 ? Best Regards, Thorsten.

Hallo, 1 und 2 lassen sich en block beantworten: die Host Software bestimmt die Aufloesung und die Geschwindigkeit, und beides ist durch das Uebertrgaungsprotokoll der Logic/Mackie Control vorgegeben. Wie so oft - auch hier kann man von der Anwenderseite nichts aendern. Nein, es ist nicht moeglich, neben den Motorfadern noch weitere Potis an einem einzigen Core zu betreiben Gruss, Thorsten.

I didn't saw this imperfection since it's already fixed in my private version. Open lc_sfb.inc, go to line 238 and change "rcall" by "call" Best Regards, Thorsten.

Hi Chriss, unfortunately the patch request function of jsynth is still not reliable, but a whol BankStick backup can be realized much easier, just follow these instructions: http://www.ucapps.de/mios/mios_backup.txt My question: has anybody already created more than 128 different patches? ;-) Best Regards, Thorsten.

I guess that you are using the vintage setup, so possibly setup_mbseq_vintage.syx? This would mean that you have to change setup_mbseq_vintage.asm, and you have to assemble this file instead of main.asm Best Regards, Thorsten.

Hi Rowan, ok, I understand this problem with the shipping cost. My concerns are due to the fact that people from europe, who could buy the encoders from Wouter, wanted to participate the collection order - and I guess that you need so much people for the minimum quantity requirements - but such actions could endanger the starting cooperation with Wouter. It's the same sensible topic like PCB mass orders (see http://www.midibox.org/cgi-bin/yabb/YaBB.cgi?board=news;action=display;num=1052767519) I don't want to forbid mass orders (I wouldn't be able to control this anyhow...), but I want to remind that they are sometimes destructive in special cases. Frank's advice is very good - search for people in the pacific area who want to order stuff from Wouter. He has some other components which could also be interesting for you (like PICs, LCDs, etc...) so in total you will possibly save more money than ordering just only the rotary encoders from China. Frank: thats ok so long Wouter doesn't offer ALPS motorfaders! ;-) Best Regards, Thorsten.

...for a small but very effective hardware change which will affect every MBHP design. Four years ago I decided to connect the SCL line of the BankStick to pin RD3 of the PIC. This was maybe the worst idea I ever had, since it required a lot of quirks in the following years to keep the firmware - and later to keep MIOS - compatible to this circumstance. This connection is the reason: why the output pins of the DOUT chain are jumping to logic-0 for about 100 uS during the SRIO update cycle (cannot be regognized by eye, but affects projects which rely on stable output levels) why LEDrings and the SID control matrix are not so bright as they could be why IIC/BankStick accesses have to block all interrupt service routines for some microseconds I never improved this part of the hardware due to compatibility reasons. I worried about the people who have to change their hardware and about my own documentation effort. Maybe the second worst idea I ever had... some years ago only 10 people had to change their hardware, in the meantime there are hundred of peoples (and who knows how much in the future... ;-)) So, it's time to align this design weakness ASAP in order to ensure best quality. The only change which has to be made is to solder J4:SC to Pin 28 instead of Pin 22. MIDIbox SID users additionally have to solder J2:SC of the SID module to J10:MD instead of J10:SC of the core module. And MIOS has to be updated, this will be a change for everybody. It would be very helpful for me if some people who already built a MIDIbox could test this change before the release. I'm especially searching for MIDIbox64E, MIDIbox SEQ and MIDIbox SID users - please contact me to get the new MIOS version (tk@midibox.org) Best Regards, Thorsten. P.S.: anybody remembers the A20 gate? http://www.win.tue.nl/~aeb/linux/kbd/A20.html I just want to prevent the same desaster ;-)

Hi Poeloq, I assume that you've missunderstood the competition - not the banners are under discussion, but the first logo made by rasOfir (see http://www.ucapps.de) However, there is a new contribution from MarcMooga -> http://logo.midibox.org Reminder: deadline is 1st May 0:00 GMT Best Regards, Thorsten.

Hallo, einfach in main.asm die Emulation ID aendern und neues .syx bauen :) ;; these settings are used in lc_init.inc to initialize the drivers #define LC_EMULATION_ID 0x14 ; use 0x10 for Logic Control ; 0x11 for Logic Control XT ; 0x14 for Mackie Control ; 0x15 for Mackie Control XT (?) Gruss, Thorsten.

Hallo, die Antwort befindet sich hier: http://www.midibox.org/cgi-bin/yabb/YaBB.cgi?board=mios;action=display;num=1082644886 Bei der Frati ist dieses Problem nicht aufgefallen, da sie mit der PIC16 Firmware arbeitet, bei der dieser Software-Workaround fuer den BankStick nicht eingebaut war. Ich werde den Workaround demnaechst durch eine minimale Hardwareaenderung ersetzen - besser ist das... Gruss, Thorsten.

Hallo, nein, Du hast 4 LED Reihen, bei denen die Kathoden nicht direkt mit Masse, sondern mit den 4 DOUT Pins verbunden werden, so dass sie im Zeitmultiplex-Verfahren geschaltet werden koennen. Und zwar so schnell, dass es Dein Auge nicht wahrnimmt. Eine Erweiterung fuer noch groessere Displayformate halte ich fuer unrealistisch. Es war schon schwierig genug, die Performance mit den 2 * 2x40 einzuhalten. Lob ist angekommen :) Best Regards, Thorsten.

Hallo, jede Kabelverbindung hat einen Widerstand, ueber den eine geringe, aber messbare Spannung abfaellt. Ausserdem wirken lange Kabelverbindungen wie eine Antenne, die Stoerungen auffaengt - die machen sich dann als Jitter bemerkar. Deshalb macht es Sinn, zumindest die Masse sternfoermig zu verdrahten. Die Vorgabe "max. 8 Potis hintereinander" ist ein Erfahrungswert. Gruss, Thorsten.

Hallo, schreib doch mal, wie Deine MIDIbox genau aussieht (anzahl der Komponenten, etc...), wie Du sie bedienen moechtest, und mit welcher Host-Software sie zusammenspielen soll. Evtl. koennte ich Dir dann auch einen konkreteren Tip geben... Gruss, Thorsten.

Hallo, spontan fallen mir zwei Fehler auf: zum einen beinhaltet der String hinter TEXT_GROUP/TEXT_BANK/TEXT_FADER/TEXT_BUTTON bereits die Cursor Position, evtl. hast Du vergessen, diese anzupassen. Zum anderen sehe ich hinter dem movf ein "BANKED" Parameter, obwohl BSR nicht initialisiert wurde - das kann zu Problemen fuehren (vor allem, da die meisten MIOS Funktionen BSR aendern). Siehe auch http://www.ucapps.de/mios/mios_ram_handling.txt Gruss, Thorsten.

Hallo, auf diese Weise lassen sich die Shift-Register direkt ohne Bitshift ansprechen - das geht schneller. Ausserdem ist es modularer (man beachte die Variationen, die bereits moeglich sind - uff!) Gruss, Thorsten.

Hi, whats the code at this line? Best Regards, Thorsten.

Hi, without loosing too much words: this is a design weakness which will be fixed in the next MIOS version. It's related to the way how additional components like the BankStick, the AOUT and the SID module are accessed from the core. Since these components are not relevant for MIDIO128, just do the following: download the MIOS source code from sourceforge, open mios_srio.inc and remove the code between line 52 and 75 (the MIOS_SRIO_Prep_Loop) Reassemble the code and convert it to a .syx file with "perl hex2syx.pl main.hex -os" This change will be part of the next release. The next release will also require to connect the SCLK line of the BankStick/AOUT/SID module to another pin. A hardware change for other modules won't be required. Best Regards, Thorsten.

You could add your "meta events" to MIDI_EVNT_Send_Fx in midi_evnt.inc Currently this function contains the code for the AOUT module, but - of course - it can be changed for any other output Best Regards, Thorsten.

Hi Robin, did you solve the problem in the meantime? Root cause: overlapping data areas. You've to remove the "org" statements and to add a single one (which points to 0x400000) to the main.asm, or you have to relocate the data blocks by hand to free addresses. Best Regards, Thorsten.